Building intelligent systems that turn data into useful decisions.
I'm a B.Tech Computer Science Engineering (AI) student at CSJM University (UIET), Kanpur. I build machine-learning systems end to end: data, model, API, deployment. My focus is applied ML and NLP for climate action and education, with an emphasis on systems that are explainable, personalised and useful outside a notebook.
name: Jitin Kumar
role: AI/ML Developer, Backend Engineer
education: B.Tech CSE (AI)
university: CSJM University (UIET), Kanpur
period: Aug 2024 - Jun 2028
cgpa: 8.71 / 10
core_stack: Python, FastAPI, Scikit-Learn
domains: ML, NLP, Data Science, Applied AI
seeking: Internships in applied AI / ML |
Currently working on
Ask me about
|
flowchart LR
A["AWS APAC Solutions<br/>Architecture<br/>Jan 2025"] --> B["ML Intern<br/>TechKriti, IIT Kanpur<br/>Jan - Feb 2025"]
B --> C["Data Science Intern<br/>Tamizhan Skills<br/>Jun - Jul 2025"]
C --> D["Green Skills and Applied AI<br/>1M1B<br/>Jun - Aug 2026"]
D --> E["Next internship"]
style D fill:#1F6FEB,color:#ffffff,stroke:#1F6FEB
style E stroke-dasharray: 5 5
Green Skills & Applied AI for Climate Action Intern | 1M1B | Jun 2026 - Aug 2026
-
Programme run with Microsoft Elevate, MeitY Startup Hub and UP Pragya.
-
Completed 70+ hours of training in AI for climate action and green skills.
-
Built and deployed SkillGreen, an ESG career-readiness prediction platform, using FastAPI, Scikit-Learn and React.
-
Built the frontend for authentication, resume workflows and skill-gap analysis, later contributed to the XERJ Corpus Hub.
-
Extending the backend with resume parsing and model explainability.
Data Science & Analytics Intern | Tamizhan Skills | Jun 2025 - Jul 2025
- Worked across the complete data workflow on 5+ datasets.
- Reduced dataset inconsistencies by about 25% through outlier detection and missing-value handling.
- Reduced data-handling time by about 40% with vectorised NumPy operations.
- Built 3 interactive Plotly dashboards for data-driven analysis.
Machine Learning Intern | TechKriti, IIT Kanpur | Jan 2025 - Feb 2025
- Built 2 supervised learning models: an NLP classifier and a regression model.
- Reduced training time by about 20% through feature selection and efficient Scikit-Learn preprocessing.
- Ranked Top 3 in the internal showcase and presented results to an IIT Kanpur faculty panel.
AWS APAC Solutions Architecture Virtual Program | Forage | Jan 2025
- Designed a scalable AWS Elastic Beanstalk hosting architecture for a simulated client requirement.
- Presented the technical recommendation clearly to a non-technical audience.
| SkillGreen | RiskLens | ModelServe |
|---|---|---|
| ESG career-readiness prediction | Insurance premium risk classifier | ML model-serving layer |
FastAPI React PostgreSQL Docker |
Scikit-Learn FastAPI |
FastAPI Jupyter |
| Repository / Live demo | Repository | Repository |
An end-to-end ML application that predicts ESG career readiness from a user's profile and provides the foundation for personalised career insights.
System architecture
flowchart LR
U(["User"]) --> F["React + Vite<br/>Vercel"]
F -->|"REST / JSON"| B["FastAPI + Pydantic<br/>Docker on Render"]
B --> M["Scikit-Learn model"]
B --> D[("PostgreSQL<br/>Render")]
M --> B
B --> F
style F fill:#1F6FEB,color:#ffffff,stroke:#1F6FEB
style B fill:#2EA043,color:#ffffff,stroke:#2EA043
Request lifecycle
sequenceDiagram
participant U as User
participant F as React Frontend
participant A as FastAPI
participant M as ML Model
participant DB as PostgreSQL
U->>F: Submit profile
F->>A: POST /predict
A->>A: Validate with Pydantic
A->>M: Engineered features
M-->>A: Readiness prediction
A->>DB: Store result
A-->>F: JSON response
F-->>U: Career-readiness insights
Build details and roadmap
- ML-based ESG career-readiness prediction with feature engineering and predictive modelling
- FastAPI + Pydantic inference backend, containerised with Docker
- React + Vite frontend on Vercel; backend and PostgreSQL database on Render
- Shipped (frontend): authentication, resume workflows, skill-gap analysis UI
- In development: backend resume parsing and model explainability
- Open source: frontend contributed to the XERJ Corpus Hub
Stack: Python Scikit-Learn FastAPI Pydantic React Vite PostgreSQL Docker
Classifies an insurance profile into Low, Medium or High premium-risk categories.
Build details
- Structured-data preprocessing with feature engineering integrated into the model pipeline
- ML-based risk classification
- FastAPI inference service exposed through a REST interface
Stack: Python Scikit-Learn FastAPI Feature Engineering
A lightweight serving layer that exposes trained models through structured REST APIs.
Build details
- Preprocessing, prediction and model-evaluation endpoints
- Auto-generated API documentation
- Jupyter-based experimentation workflow
Stack: Python FastAPI REST APIs Jupyter
Combines academic and behavioural signals to identify students who may need early intervention.
Build details
- Combined grades, attendance and login activity into one dataset
- Random Forest classifier reaching 87% accuracy in identifying low-performing students
- Dashboard-oriented data pipeline with per-student improvement suggestions
Stack: Python Pandas Scikit-Learn Matplotlib Jupyter
Contributed the SkillGreen React frontend to the XERJ Corpus Hub.
- 25 files, 5,700+ lines of code
- Covers authentication, resume workflows, skill-gap analysis and the user interface
- Contributor profile
Stack: React Vite JavaScript
flowchart LR
A["Data<br/>cleaning, outliers,<br/>missing values"] --> B["Features<br/>engineering and<br/>selection"]
B --> C["Model<br/>training and<br/>evaluation"]
C --> D["API<br/>FastAPI +<br/>Pydantic"]
D --> E["Deploy<br/>Docker, Render,<br/>Vercel"]
style A fill:#161B22,color:#ffffff,stroke:#30363D
style B fill:#161B22,color:#ffffff,stroke:#30363D
style C fill:#1F6FEB,color:#ffffff,stroke:#1F6FEB
style D fill:#161B22,color:#ffffff,stroke:#30363D
style E fill:#2EA043,color:#ffffff,stroke:#2EA043
mindmap
root((Jitin Kumar))
Machine Learning
Predictive Modelling
Feature Engineering
Random Forest
Linear Regression
NLP
Model Evaluation
Backend
FastAPI
Pydantic
REST APIs
PostgreSQL
Frontend
React
Vite
Deployment
Docker
AWS
Vercel
Render
Data
NumPy
Pandas
Plotly
| Certification / Activity | Organization |
|---|---|
| Machine Learning Certification | TechKriti, IIT Kanpur |
| Data Science & Analytics Certification | Tamizhan Skills |
| Machine Learning with Python | freeCodeCamp |
| Green Skills & Applied AI for Climate Action | 1M1B |
| Open Source Contributor | XERJ Corpus Hub |
| National Service Scheme | NSS Volunteer |